{"id":"W4407174976","doi":"10.3389/frai.2025.1533508","title":"Large language models generating synthetic clinical datasets: a feasibility and comparative analysis with real-world perioperative data","year":2025,"lang":"en","type":"article","venue":"Frontiers in Artificial Intelligence","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":24,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"","keywords":"Computer science; Fidelity; Data mining; Sample (material); Missing data; Sample size determination; Artificial intelligence; Machine learning; Statistics; Mathematics","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001834617,0.0002536612,0.0008368334,0.0005944066,0.0002656446,0.0001069775,0.0003403026,0.000131353,0.00005991958],"category_scores_gemma":[0.0004555115,0.0002146812,0.00007494992,0.002046816,0.000465037,0.0003855375,0.0001937764,0.0005380045,0.000008445939],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002183854,"about_ca_system_score_gemma":0.000455991,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002435978,"about_ca_topic_score_gemma":0.02942317,"domain_scores_codex":[0.996673,0.0003936937,0.001180365,0.001060051,0.0002568912,0.0004360384],"domain_scores_gemma":[0.9978191,0.0003761169,0.0001713688,0.001254917,0.0001932191,0.0001852518],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002631147,0.002142603,0.7692522,0.0002692706,0.001303297,0.0001047438,0.02803481,0.007225737,0.0001409222,0.02642079,0.003330748,0.1591437],"study_design_scores_gemma":[0.00006173653,0.0002140044,0.00487,0.0002550975,0.000686706,0.000003071379,0.05984274,0.9278378,0.002556947,0.00290391,0.0004670477,0.000300903],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5639937,0.001588276,0.4295726,0.001010965,0.000564301,0.00114039,0.0003763746,0.00006021413,0.001693196],"genre_scores_gemma":[0.9720501,0.0004740462,0.02583555,0.0004703086,0.0001269013,0.00004435989,0.000760544,0.000009414623,0.000228723],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9206121,"threshold_uncertainty_score":0.9882873,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3388162493346323,"score_gpt":0.5062546044831657,"score_spread":0.1674383551485334,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}